Water pressure


Published on

  • Be the first to comment

  • Be the first to like this

Water pressure

  1. 1. SECTION 9.21 WATER PRESSURE BOOSTER SYSTEMS SATORU SHIKASHOTYPES OF WATER PRESSURE BOOSTER SYSTEMS _______________________Plumbing fixtures and equipment connected to a water source will function properly onlywhen a minimum water supply pressure is consistently available. Whenever this mini-mum pressure cannot be maintained by the supply source, a means of boosting the waterpressure should be considered. The commonly used methods for boosting pressure are 1. Elevated gravity (tank) systems 2. Hydropneumatic tank systems 3. Variable-speed-drive centrifugal pump booster systems 4. Tankless constant-speed multiple centrifugal pump systems 5. Limited-storage, constant-speed multiple centrifugal pump systems Table 1 identifies the significant advantages and disadvantages of each system. Although gravity and hydropneumatic tank systems have some distinct advantages,most installations favor constant-speed multiple-pump and variable-speed-drive systems.Because gravity and hydropneumatic tank systems are now seldom specified, they will notbe discussed in detail here. Multiple-pump systems are well suited for high-rise buildings,apartment buildings, schools, and commercial installations. Variable-speed-drive systemsare particularly suited for industrial applications, where control precision is required andmaintenance capability for complex electronic apparatus is present. It is important to remember that the primary function of a pressure booster pump fora building, whether variable- or constant-speed, is to maintain the desired system pres-sure over the entire design flow range. It is also important to recognize that the cold waterservice piping configuration is an open-loop system, not a closed-loop system as it is in 9.447
  2. 2. 9.448 CHAPTER NINETABLE 1 Comparison of significant factors for pressure booster systemsSystem type Advantages DisadvantagesElevated gravity Simplicity, large storage Size and weight, water damage tank capacity, low energy usage, potential, freeze potential if reserve storage capacity for roof-mounted, corrosion and fire protection contamination potential, limited pressure for floor immediately below tank, possible unsightly appearance, periodic cleaning and paintingHydropneumatic Location not critical, low energy Requires compressed air tank usage, limited storage source, corrosion and capacity, pump does not run contamination potential, when there is no demand, large pressure variation, operation in optimum pump relatively large, standby flow range provision costlyVariable-speed Simple controls, off-the-shelf Slow response to sudden drive: Fluid motor, standby provision demand change, may require couplings less costly than for above heat exchanger to cool drive, units slip losses result in lower maximum speed and higher motor power, requires selective application, no water storageVariable-speed Low motor current inrush, Complex electric circuitry, high drive: ac type precise pressure control, initial cost for low-power higher than 3600 rpm units, may require special possible, few mechanical motors, motor low-speed devices, large power limitation, rapidly changing capability at lower first cost technology, requires selective applicationTankless Relatively low first cost, uses Continuously running lead constant-speed time-proven components, pump, difficult to accurately multiple pump compact size, inherent partial determine capacity split standby capacity, extra among pumps, no water standby capacity inexpensive, storage, problems associated location not critical, good with low flow rates pressure regulationLimited-storage Shuts down during very low No significant disadvantages, constant-speed water demand, uses time- tanks with high maximum multiple pump proven components, standby working pressure may be capacity inexpensive, location difficult to obtain not critical, limited water storage, no air-to-water contamination with diaphragm tankheating and cooling systems. With an open-loop system, the static head above the pump isnot canceled by the down leg of the piping, and therefore the pump must develop a headequal to or greater than the static head, even at zero flow (with suction pressure atdesign). Review Sections 8.1 and 8.2 for further discussions on closed-loop systems.
  3. 3. 9.21 WATER PRESSURE BOOSTER SYSTEMS 9.449Variable-Speed-Drive Pressure Booster Systems Initial variable-speed drives werefluid couplings, magnetic couplings, and liquid rheostat (wound rotor motor) drives. Thecoupling-type drives were driven by a constant-speed motor, with the coupling output shaftvarying in speed. With the advent of solid-state electronics technology and circuit minia-turization, most variable-speed drives currently used for pressure booster applications areof the solid-state, ac, adjustable-speed type. These drives are discussed in Subsection 6.2.2. Variable-speed drives are usually specified for their low operating cost potential. Toachieve the energy-savings goal, the system conditions should cause the drive speed tovary between 50 and 75% of full speed during most of the operating period. The pump speed, head, and flow relationships are expressed by the affinity laws:• Flow varies directly as speed: Q2 N2 ϭ (1) Q1 N1• Head varies directly as the square of the speed: ϭ a b H2 N2 2 (2) H1 N1where Q ϭ flow rate, gpm (m3/h) N ϭ pump speed, rpm H ϭ pump total head, ft (m) The required pump total head at design conditions is H ϭ 1Hd Ϫ Hs 2 ϩ Hf (3)where H ϭ required pump total head at design conditions, ft (m) Hd ϭ design system pressure at point of control, ft (m) Hs ϭ minimum design suction pressure, ft (m) Hf ϭ sum of all losses between Hd and Hs at design flow, ft (m) At flow rates less than design conditions and suction pressures higher than the mini-mum design pressure, the pump head requirement is reduced by the change in pipe fric-tional losses Hf and the additional suction pressure Hs available. Also, because of the risingcharacteristic of the centrifugal pump performance curve as the flow decreases, an excesshead is developed by the pump. All of these changes will alter the required pump totalhead and the pump speed from the design (maximum) values. For a single operating pump unit, design maximum speed will be required at designminimum suction head and design maximum flow. Minimum speed will be required atmaximum suction head and minimum flow. The procedure required to determine thespeed range of the pump driver requires construction of the system-head curve, the pumphead-capacity curve (see Section 8.1), and the affinity curve. Rearranging the affinityequations (Eqs. 1 and 2), the speed changes are calculated: N2 ϭ N1 a b Q2 (4) Q1 N2 ϭ N1 a b H2 1>2 (5) H1where the subscripts 1 and 2 represent the higher and lower speed values, respectively, forpump conditions at the same specific speed (see Subsection 2.3.1) or along the same affin-ity line.
  4. 4. 9.450 CHAPTER NINE The following examples illustrate how changes in flow rate, pipe frictional losses, andsuction head affect pump speed. EXAMPLE 1 With no change in suction pressure—first (design) conditions: Hd ϭ 193 ft (58.8 m) Hs ϭ 50 ft (15.2 m) Hf ϭ 7 ft (2.1 m) Q ϭ 190 gpm (43.1 m3/h) N ϭ 3500 rpm H ϭ (193 Ϫ 50) ϩ 7 ϭ 150 ft (45.7 m) (Eq. 3) Second conditions: Hd ϭ 193 ft (58.8 m) Hs ϭ 50 ft (15.2 m) Hf ϭ 1.9 ft (0.58 m) Q ϭ 100 gpm (22.7 m3/h) N ϭ to be calculated H ϭ (193 Ϫ 50) ϩ 1.9 ϭ 144.9 ft (44.18 m) (Eq. 3) The 3500-rpm pump head-capacity curve, the system-head curves, and the affinity (square) curves are shown in Figure 1. The operating points are lettered. Point A is for maximum flow at minimum suction head (design conditions). Point B is for lower flow at minimum suction head at a speed to be determined. The reduced piping losses between head points Hd and Hs are represented by the system-head curves. To determine the approximate speed at point B, it is necessary to use trial and error because the flow and head ratios in Eqs. 4 and 5 are not known. The following proce- dure may be used to estimate the speed (Figure 1): 1. Draw an affinity (square) curve passing through the zero-flow/zero-head point and the lower operating point (point B) and intersecting at head-capacity curve of known speed (point C).FIGURE 1 System and pump curves to determine speed reduction for the 2-in (51-mm) pump in Examples 1 and 2. (ft ϫ 0.3048 m; gpm ϫ 0.2271 = m3/h)
  5. 5. 9.21 WATER PRESSURE BOOSTER SYSTEMS 9.451 2. Read the probable flow rate at point C, the point of intersection between the affinity curve and the pump curve of known speed. 3. Calculate the head relative to the flow rate read at point C, based on the square curve relationship: H2 H1 ϭ (6) a b N2 2 N1 where subscripts 1 and 2 represent the values at the 3500-rpm and lower speeds (points C & B), respectively. 4. Compare the head calculated in step 3 with the head of the pump curve (point C) of known speed at the probable flow rate of step 2. If the values differ significantly, try another flow rate and repeat steps 3 and 4. 5. Using the accepted head from step 4, calculate the speed at the operating point (point B) using Eq. 4 or 5. The equation not used may then be used for verification. For this example, the estimated flow rate at point A (step 2) is 104 gpm (23.6 m3/h) andthe calculated head at the same point (steps 3 and 4) is 144.9 44.18 H1 ϭ ϭ 156.7 ft or ϭ 47.76 m a b a b 100 2 22.7 2 104 23.6 The speed at the operating point (point B) is N2 ϭ 3500 a b 3500 a b ϭ 3365 rpm 100 22.7 or (Eq. 4) 104 23.6 Check: N2 ϭ 3500 a b 3500 a b ϭ 3366 rpm 144.9 1>2 44.17 1>2 or (Eq. 5) 156.7 47.76 The speed change is 3500 Ϫ 3365 ϫ 100 ϭ 3.9% 3500 EXAMPLE 2 With change in suction pressure—first conditions: same as Example 1. Second conditions: same as Example 1, except Hs ϭ 110 ft 133.5 m2 H ϭ 1193 Ϫ 1102 ϩ 1.9 ϭ 84.9 ft 125.88 m2 1Eq. 32 Operating point B¿ and other curves related to this example are shown in Figure 1. Again by trial and error, the flow rate and pump head at the intersection of the affin- ity curve and the 3500-rpm pump curve (point A¿) are 134 gpm (30.4 m3/h) and 152.4 ft (46.45 m). The speed at the operating point is N2 ϭ 3500 a b 3500 a b ϭ 2612 rpm 100 22.7 or (Eq. 4) 134 30.4
  6. 6. 9.452 CHAPTER NINE Check: N2 ϭ 3500 a b 3500 a b ϭ 2612 rpm 84.9 1>2 25.88 1>2 or (Eq. 5) 152.4 46.45 The speed change is 3500 Ϫ 2612 ϫ 100 ϭ 25.4% 3500 These examples indicate that the speed of the pump is not significantly affected unlessthere is an appreciable change in suction pressure. For optimum performance, it should be confirmed, after the speed has been deter-mined, that the most often occurring flow demand range is ideally near the pump’s maxi-mum efficiency. Generally, the design point at maximum speed should be selected to the right of thepump’s best efficiency point. The operating flow range at different speeds should be withinthe hydraulically and mechanically stable ranges of the pump. The pump shaft power at any specified speed is QH1sp. gr.2in USCS units hp ϭ 3960E QH1sp. gr.2in SI units kW ϭ (7) 367Ewhere H ϭ pump total head, ft (m) Q ϭ flow rate, gpm (m3/h) sp. gr. ϭ specific gravity of fluid E ϭ pump efficiency, % (expressed in decimal equivalent) H, Q, and E are values from the pump performance curve for the specified speed andimpeller size. The motor power is the same as the shaft power for pumps driven directly by the motor.For pumps driven through intermediate variable-speed couplings, the motor power mustalso include the drive slip and fixed loses. When the required design flow exceeds the capacity of a single pump, several pumps,including possibly one constant-speed unit, can be arranged to operate in parallel. Twomethods of staging the pumps are usually used: 1. The first pump is operated in the variable-speed mode until its maximum speed is reached; the second pump, also a variable-speed unit, is energized. Now both pumps are operating in parallel at the same reduced speed. This sequence is repeated for the other pumps. 2. The first pump is operated in the variable-speed mode until its maximum speed is reached and is then locked in at this speed to operate as a constant-speed pump. The second pump is energized and operates in the variable-speed mode. The sequence is repeated for the other pumps. The speed variation of the second pump is relatively small because it must develop the same head as the first pump operating at maximum speed. Because the flow rate through the second pump is less, its speed is affected by the increase in total head available from the rise in the pump performance curve. Which method of sequencing is selected depends on economics, equipment redundancy,and other considerations. For example, with the first method, both pumps must be fur-nished with variable-speed drive units. With the second method, only one variable-speeddrive unit is required if it is an electronic type because the first pump is locked in by sep-arate electric means.
  7. 7. 9.21 WATER PRESSURE BOOSTER SYSTEMS 9.453 The sole function of the pumps is to maintain constant pressure at the control point;therefore, the controller selected must be pressure-actuated. The type (follower signal) ofcontroller chosen depends on the type of speed change signal acceptable by the variable-speed control unit. For electronic motor speed controls, such as variable-voltage andvariable-frequency units, typical follower signals are low-voltage dc, milliamp dc, 135-ohmpotentiometer, and pneumatic.Tankless Constant-Speed Multiple-Pump System The major components of this sys-tem (Figure 2) areFIGURE 2 Constant-speed multiple-pump pressure booster systems: (a) horizontal end-suction volute pumps(ITT Bell and Gossett), (b) vertical end-suction volute pumps (ITT Bell and Gossett), (c) vertical turbine pumps withlimited storage (SynchroFlo)
  8. 8. 9.454 CHAPTER NINE FIGURE 3 Constant-speed multiple-pump pressure booster schematic (ITT Bell and Gossett)FIGURE 4 Typical pressure-reducing and check valve to maintain constant system pressure and preventbackflow (Cla-Val) 1. One or more pumps, with two or three most common 2. A combination pressure-reducing and check valve (PRCV) for each pump (parallel- piped PRCVs are used with larger sizes; separate pressure-reducing valves and check valves may be used) 3. An automatic sequencing control panel 4. When factory assembled, a steel frame for the entire unitPIPING ARRANGEMENT AND FLOW PATH A schematic of the piping arrangement and flow pathin a constant-speed multiple-pump system is shown in Figure 3. Supply water under fluc-tuating pressure enters the suction header and flows into the pump, where it is boosted toa higher pressure. This varying high-pressure water enters the PRCV (Figure 4), and thepressure is reduced to the constant pressure desired over the design flow range. Flow rever-sals through the idle and parallel pump circuits are prevented by the checking feature ofthe PRCV, which also dampens the pressure fluctuations caused by sudden flow changes. When only minor changes in supply water pressure are anticipated, such as from anonpressurized tank, and the pump head-capacity curve is relatively flat, silent checkvalves without pressure-reducing feature may be used. This will result in a slight increasein discharge pressure above the desired design constant pressure. When pumps of differ-ent sizes are used, care should be exercised to avoid having the higher-head pump force
  9. 9. 9.21 WATER PRESSURE BOOSTER SYSTEMS 9.455the lower-head pump to shut off or to operate at less than minimum design pump flow. Adecrease in flow from a centrifugal pump must be accompanied by an increase in pumptotal head as required by the pump head-capacity curve. If check valves only are used, itis preferable to have all pumps and valves identical to avoid unbalanced flows. Operating the pumps at shutoff (no flow) will cause the water temperature in the pumpcastings to rise (Subsection 2.3.1). To keep the temperature in the pumps within a safe limit,the heated water is relieved through the thermal relief valve. This valve may be either a self-actuated (thermostatic) type or a solenoid valve actuated by a temperature controller.PUMP CONTROL PANEL One of the advantages of a factory-assembled pressure boosterpackage is the prewired, pretested, pretubed, mounted control panel that requires a min-imum of field connections. In addition to providing for the proper sequencing of the pumps,the control panel should contain electrical interlocks for the operating and safety controlsand circuit connections for remote control units. Standard items and optional equipment vary considerably from one manufacturer toanother. The components usually included with a standard panel are listed: Steel enclosure Starters Control transformer Sequencing controllers Control circuit protector Pump failure interlocks Selector switches Minimum-run timers Low-suction pressure control Time delays Pilot lights Optional features that may be available: Power supply fused disconnects or circuit breakers Low system pressure control Enclosure door interlock Low water level control High water temperature control Emergency power switchover High system pressure control Unit failure alarm Pump alternation Low-flow shutdown Program time switch Miscellaneous enclosure types Elapsed time meters Power economizer circuit Additional pilot lights Most factory-wired panels conform to one or more of the consumer safety agencies,such as Underwriters’ Laboratories (UL), National Electrical Code (NFPA/NEC), andCanadian Standards Association (CSA) and are furnished with a label so indicating.PUMP CONTROL SEQUENCE A typical elementary wiring diagram provides the followingsequences of events. With both pump selector switches in the auto mode and all safetyand operating controls in run status, the lead pump starter is energized, starting pump1. As the system water demand increases, a staging control switch, which senses motorcurrent, flow, or system pressure, starts pump 2. Pump 2 continues to operate until thedecreasing water demand causes the staging control switch to open, stopping pump 2. Ifthe circuit is provided with a minimum-run timer, pump 2 will continue to run for the settime period regardless of the staging switch status. This timer prevents pump 2 fromshort-cycling during rapidly fluctuating demand periods. For test purposes and emergency operation, both pumps may be operated by placingthe selector switches in the hand mode. In this position, most of the safety and operatingcontrols are bypassed. Should pump starter 1 fail to operate because of an overload heater relay trip or startermalfunction, a failure interlock switch automatically starts pump 2. Pump 1 will run continuously unless the circuit is provided with shutdown features,such as high-suction pressure control and/or low-flow shutdown control.
  10. 10. 9.456 CHAPTER NINE Many units are specified for manual or automatic alternation of the equally sizedpumps. This feature is intended to equalize the wear among the pumps and associatedcomponents. Alternation of any pump designated for standby duty (emergency use) maynot be prudent. The standby pump should be preserved until needed, similar to an emer-gency power generator.Limited-Storage Constant-Speed Multiple-Pump System The most notable short-coming of the tankless multiple-pump system is the need for a continuously runningpump, even at zero water usage. This drawback is remedied by the addition of a pressur-ized storage tank connected to the high-pressure side of the piping system with low-flowshutdown controls to stop the pumps. The limited-storage system is not a scaled-down version of the hydropneumatic sys-tem. It differs in three important aspects: 1. The primary function of the tank is water storage. 2. Tank pressure is developed by the booster pump. 3. Air and water in the tank are separated by a flexible barrier (diaphragm, bladder) to eliminate direct interaction between the gas and water in the tank. The barrier also prevents the gas from escaping when the tank is emptied of water.SEQUENCE OF OPERATION CONTROL CIRCUITRY During periods of normal water usage, theoperating sequence is that of the tankless unit. As the water flow approaches zero, a low-flow sensing device stops the pumps. The tank provides the water needs during the shut-down period until the water pressure diminishes to the minimum allowable value. At thistime, a pressure switch starts the lead pump to restore the pressure and recharge thetank with water. The low-flow device may be a flow switch, a pressure switch, or a temperature-actuatedswitch and must be capable of switching at flows below 5 gpm (1.14 m3/h). The tank should be sized for sufficient drawdown volume (water available from tankduring shutdown) to avoid excessive pump cycling.RELATIONSHIPS BETWEEN TANK SIZE, DRAWDOWN VOLUME, AND PRESSURE The approximate for-mula for sizing prepressurized diaphragm tanks at constant temperature is Ve Vt ϭ (8) Pf 1Ϫ P0where Vt ϭ tank volume, ft3 (m3) Ve ϭ change of gas volume in tank, ft3 (m3) Pf ϭ final gas pressure lb/in2 (kPa) abs P0 ϭ initial gas pressure, lb/in2 (kPa) abs The corresponding relationships of water volume and pressures in the tank are• Ve is equal to the drawdown volume.• Pf is the water pressure at the end of the drawdown cycle, the minimum allowable sys- tem pressure.• P0 is the water pressure at the beginning of the drawdown cycle or at the termination of the charging cycle.TANK LOCATION Because both Pf and P0 are influenced by the static head above the tankand the available charging pressure, the location of the tank and point of connection tothe system should be carefully selected. Connecting the tank at A in Figure 5 providesthe highest available tank charging pressure because this pressure is not affected by thereduction through the PRVs. However, there are several disadvantages to this point ofconnection:
  11. 11. 9.21 WATER PRESSURE BOOSTER SYSTEMS 9.457FIGURE 5 Relative effect of tank location on tank size and pressure for equal draw-down volumes (ITT Bell andGossett) 1. The highest static pressure is applied to the tank because the pumps are usually located on the lowest floor of the building. 2. The tank may be subjected to high working pressure because any increase in suction pressure above the minimum design pressure is additive to the pump head. 3. The tank and all alternated lead pumps must be interconnected. 4. A silent check valve must be placed between the tank connection point and the pump discharge nozzle. Connecting the tank in B in Figure 5, downstream of the PRVs, will eliminate disad-vantages 2, 3, and 4. The available charging pressure is less, and therefore a larger tankis required. However, tank location is not critical. Locating the tank on the upper elevation of the building, such as at C, reduces the sta-tic head pressure on it; consequently, a smaller tank will suffice. The maximum designworking pressure is also lower, and therefore it may be possible to use a tank with a lowerpressure rating at less cost.PRESSURE BOOSTER SYSTEM SIZING__________________________________Design Data The proper sizing of a pressure booster system is subject to the accuracy ofthe design data available. Experience has indicated that the greatest single cause for unsat-isfactory pressure booster performance is oversizing. Submitted data may be rough esti-mates or historical data not relevant to the location; therefore, their source and accuracy
  12. 12. 9.458 CHAPTER NINEshould be scrutinized. Pertinent factors are the flow demand rate and hourly profile, pres-sure conditions, and types of building occupancy.FLOW DEMAND RATE AND HOURLY PROFILE Estimating water demand rate is one of the mostdifficult and controversial subjects. To determine the water demand, many designers referto the classical Hunter curve (National Bureau of Standards, Report BMS79, by R. B.Hunter) or a modified version of it for lack of a more accurate method. Numerous studieshave indicated that the Hunter method often results in very appreciable demand rateoverstatement. It should be noted that the Hunter curve was intended for determiningpipe sizes, not water demand. The average hourly demand profile is helpful in proportioning the system demand rateamong the pumps in the system. Optimum selection results in the lowest total energy con-sumption by the pumps while maintaining the demand requirements.PRESSURE CONDITIONS Four accurate pressure values are required: design system pres-sure, minimum suction pressure, maximum suction pressure, and minimum allowablesystem pressure. The difference between design system pressure and minimum suctionpressure is used to determine the developed head of the pump at design flow. Inaccuratevalues will result in incorrect impeller sizing and incorrect motor power selection. The design system pressure can be calculated with reasonable accuracy. However,when the water supply is from the municipal water main, obtaining accurate values of suc-tion pressures for the proposed site of installation is very difficult because of the absenceof recent pressure data. Often the minimum and maximum pressures are haphazardlyestimated from average citywide values. A minimum suction pressure error as small as5 lb/in2 (35 kPa) could result in selection of the next larger motor. The maximum suction pressure will determine the working pressures required of thepumps and piping and whether there is sufficient pressure to justify a bypass connection(Figure 3) paralleling the pumps. Knowing the minimum allowable system pressure is helpful in setting controls associ-ated with this pressure.TYPE OF OCCUPANCY If the hourly demand profile is not available, the type of building occu-pancy could be used to determine the standby capacity requirement, whether low-flowshutdown is applicable, and the number of pumps for the system. Most pressure boostermanufacturers provide data for determining demand, capacity splits, and pump sizing.Pump Sizing Procedures Information should be collected on system information andcomponent performance. The system information required is Qd ϭ total design capacity, gpm (m3/h) Pd ϭ design pressure at system header outlet, lb/in2 (kPa) Ps ϭ minimum available pressure at suction header inlet, lb/in2 (kPa)Q1, Q2, Q3 ϭ design capacity of individual pumps, gpm (m3/h) Components performance data required are• Pump performance characteristic curve• Pressure-reducing valve (and check valve) flow chart* (Figure 6) Next, the required pump total head at the maximum flow design condition should becalculated: H ϭ 1Hd Ϫ Hs 2 ϩ Hv ϩ Hu (9)where H ϭ required total head, ft (m) Hd ϭ design system pressure at PRCV outlet, ft (m) *Add loss through separate check valve if not integral with PRV.
  13. 13. 9.21 WATER PRESSURE BOOSTER SYSTEMS 9.459 FIGURE 6 Typical PRV or PRCV flow chart (ft ϫ 0.3048 = m; gpm ϫ 0.2271 = m3/h; in ϫ 25.4 = mm) (Cla-Val) Hs ϭ minimum design suction pressure, ft (m) Hv ϭ PRCV pressure loss, ft (m)* Hu ϭ sum of all unaccounted losses in the pressure booster package, ft (m) For pressure booster applications, the PRCV is usually sized for a nominal pipe veloc-ity between 8 and 18 ft/s (2.4 and 5.5 m/s). The pressure loss Hv is based on the valve being80% open (Figure 6). Operating the valve at less than full-open position is desirable forgood pressure regulation. The term Hu represents the allowance for piping frictional lossesbetween the suction header inlet and the system header outlet at design capacity, exclu-sive of the PRCV loss. Manufacturers of booster packages differ in determining this value.Some ignore it completely to compensate for usual oversizing as the result of inaccuraciesin the design capacity and minimum suction pressure statements. Others assign a fixedvalue, about 5 ft (1.5 m), or a percentage, about 3%, of the difference of system and suctionpressures (Hd Ϫ Hs). It is recommended that some value be used. Then determine the best combination of pump and PRCV sizes. The selection willdepend on whether the design criterion is least capital cost or lowest operating cost. Usu-ally the smallest pump and PRCV combination will result in lowest first cost and thepump equipped with the smallest motor will yield the lowest operating cost. Variouspump, PRCV, and motor combinations may be examined by plotting system curves (Figure7) for two or three PRCV sizes on one or more pump performance curves. The pumps tentatively selected should exhibit the highest efficiency at the most fre-quent operating flow point on their head-capacity curves. When pumps are operated atsuction pressures higher than design minimum, the PRCVs will throttle the excess suctionhead and maintain the same pump total head and constant design discharge pressure.When the flow demand is less than the design maximum, characteristically, the head onthe centrifugal pumps will increase to some higher value on the pump head-capacitycurve. The PRCVs will reduce this excess head to maintain the constant design dischargepressure. Therefore, each pump operates along its own head-capacity curve, which is alsothe adjusted system-head curve. Add loss through separate check valve if not integral with PRV.
  14. 14. 9.460 CHAPTER NINEFIGURE 7 System curves for PRCV sizing (ft ϫ 0.3048 = m; gpm ϫ 0.2271 = m3/h; hp ϫ 0.7457 = kW; in ϫ 25.4= mm) (ITT Bell and Gossett) When several pumps are to be used, especially pumps of different sizes, their head-capacity characteristics should be compatible. A pump having a shutoff total head lessthan the system total head at any operating point (with or without other pumps operat-ing) will not pump. Also, pumps should be so selected and controls should be such that noone pump will operate at too low a flow; that is, at a flow less than the minimum recom-mended for mechanical and hydraulic stability. If the minimum suction pressure is verylow, the net positive suction head required by the pumps throughout their operating flowrange must be checked (Subsection 2.3.1).PUMP AND PRCV SELECTION EXAMPLE Total system design capacity Qd ϭ 380 gpm (86.3 m3/h) Design capacity Q1, Q2 ϭ 190 gpm (43.1 m3/h) Design pressure at system header outlet Hd ϭ 210 ft (64.0 m) Minimum pressure at suction header Hs ϭ 70 ft at 100 ft NPSH inlet (21.3 m at 30.5 m NPSH) PRCV loss, Figure 6 Hv ϭ 36 ft (11 m) for trial 2-in (51-mm) valve Allowance for internal frictional losses Hu ϭ 5 ft (1.5m) Calculate H: In USCS units H ϭ (210 Ϫ 70) ϩ 36 ϩ 5 ϭ 181 ft* (Eq. 9) In SI units H ϭ (64.0 Ϫ 21.3) ϩ 11 ϩ 1.5 ϭ 55.2 m Table 2, calculate and tabulate H for all applicable PRCV sizes up to the design flowrate of 190 gpm (43.1 m3/h). Hv and Hu terms vary as the square of the flow rate. Now locate H at 190 gpm (43.1 m3/h) on the applicable pump curve. Select the pumpand PRCV combination that best satisfies the design criterion of least first cost or lowestenergy input. *Due to the many approximations, decimal amounts are not used in the final results.
  15. 15. 9.21 WATER PRESSURE BOOSTER SYSTEMS 9.461TABLE 2 Required head for pressure reducing valves Flow, gpm (m3/h)PRV size, in (mm) 0 (0) 50 (11) 100 (23) 150 (34) 190 (43)2 (51) 140 (42.7) 143 (43.6) 151 (46.0) 166 (50.6) 181 (55.2)21 (64) 2 140 (42.7) 141 (43.0) 146 (44.5) 153 (46.6) 160 (48.8)3 (76) 140 (42.7) 141 (43.0) 143 (43.6) 147 (44.8) 151 (46.0) Note: All values are head in ft (m). Plot H values on the selected pump curve to obtain the system curves for the threePRCV sizes (Figure 7). The plots reveal that the 3-in (76-mm) valve, a 10-hp (7.5-kW)motor, and 63 -in diameter (162-mm) impeller are the proper selections. 8 Note that if the design capacity of the pump could be reduced to about 178 gpm(40.4 m /h, a 21 -in (64-mm) PRCV with the impeller sized for 61 -in (165-mm) diameter 3 2 2may be used without additional power input. This choice reduces the cost and improvesthe PRCV performance at low flow rates. Using a 2-in (51-mm) PRCV would require a 15-hp (11-kW) motor operating at about12.5 hp (9.3 kW) load. This choice is not energy effective. The power required to drive the pump may be approximated by using the pump curveand Eq. 7. The available NPSH is in excess of the required pump NPSH (Figure 7).PUMP TYPES AND MATERIALS_________________________________________Single-stage volute centrifugal pumps, followed by multistage vertical turbine diffuserpumps in a suction tank, are most commonly used for domestic water pressure booster sys-tems. The volute pumps can be end-suction with a vertically split case, double-suction withan axially split case, or in-line. For high-pressure service, two-stage axially split casepumps or multistage vertical turbine pumps may be selected. In selecting the type of pump best suited for the application, such inherent character-istics as low-flow recirculation, low-flow cavitation, high-flow cavitation, steepness of thepump curve, noise, and operating efficiencies should be evaluated. These factors are dis-cussed in Subsection 2.3.1. The most often discussed factor in booster application is low-flow recirculation. Lowpump flow conditions do occur; they cannot be completely designed out of the system. Mostwell-designed small pumps with positive (above atmospheric) suction pressure can oper-ate safely in the low-flow region, with the only point of concern being water temperaturerise at or near shutoff flow. Should radial thrust be a major consideration, double-volutepumps or diffuser (vertical turbine) pumps offer advantages. Vertical pumps generally require less space. Pumps may be close-coupled to their dri-vers, or, if not, a coupling requiring careful alignment is required. Close-coupled pumpsmay require less maintenance, as there are no pump bearings. Shaft sealing may be eitherpacking or mechanical seal. All pumps and drivers in packaged units are mounted on a common steel frame, butindividual bases for each pump and driver may be necessary for larger units. The materials of construction for pumps, valves, and piping should be suitable for thewater quality and conditions furnished to the system. Because water impurities vary fromlocation to location, careful analysis is recommended. Federal and local agencies, such asthe Food and Drug Administration, may also prescribe allowable materials.FURTHER READING __________________________________________________American Society of Heating, Refrigerating, and Air-Conditioning Engineers. EquipmentHandbook, ASHRAE, Atlanta, GA, 1983.
  16. 16. 9.462 CHAPTER NINEITT Bell and Gossett. “Domestic Water Service.” Bulletin TEH-1175, Morton Grove, IL, 1970.ITT Fluid Handling Division. “Pressurized Expansion Tank Sizing/Installation.” BulletinTEH981, Morton Grove, IL, 1981.Potter, P. J. Steam Power Plants, Ronald Press, New York, 1949.Steel, A. High Rise Plumbing Design, Miramar, Los Angeles, 1975.